[Immunization] How to close Tasks based on the selected items on the Record All forms | Code or Content Changes

[dubdabasoduba]

• We have 2 options to achieve this.

1. Content Changes
   • The Questionnaire.item.answerOption.valueCoding.code is the Task.id for the Tasks represented by
     the Questionnaire.item.answerOption.
     • Write a structure map function to update Task.status to completed.
     • Use the Questionnaire.item.answerOption.valueCoding.code as the Task.id.value.

• Advantages
  • We don't have to add new code to process the Tasks status.
• Disadvantages
  • This behaves very differently from the Tasks closure for a form launched by specific tasks. This increases the
    maintainability complexity
  • The tasking workflow is very key for the FHIR Core functionality. Letting the content authors determine when the
    flow is complete is a risk.

2. Code Changes
   • Update the ActionParameters to add a new parameter to define the Questionnaire Item that holds displays the
     CheckBox showing the Tasks.
     • This parameter should be a string containing the Questionnaire.item.linkId
   • Write a function that iterates through the QuestionnaireResponse to fetch the answers with
     a QuestionnaireResponse.item.linkId that matches the definition on the ActionParameters.
     • Get the QuestionnaireResponse.item.answer.valueCoding.code from each answer.
     • Use the list of QuestionnaireResponse.item.answer.valueCoding.code as Task IDs to fetch Tasks from the
       SQLite database.
     • Update all the returned Task.status to completed

• Advantages
  • We maintain the same architecture where the code handles closing the loop on the tasking workflow.
  • The users don't have to worry about determining when tasks are completed.
• Disadvantages
  • We will be adding extra code targeting Tasks status updates.

[pld]

I prefer the content approach, at least at this point. It's hard to know yet what is going to be the best way to handle tasking workflows, if there was an obvious general approach, or one emerges, then I'd agree we should consider hard-coding that, but I don't think we have the data to determine that yet